Biology 202, Spring 2005
Second Web Papers
On Serendip

First of all, the illusions evident in many works of art should not be thought of as "optical illusions" in the sense of images that are meant to trick the brain into seeing something that isn't there. Instead, they should be thought of as what Patrick Cavanagh calls a "simplified physics" employed by our brains to interpret 2D works of art into a 3D image in our minds (1). This simplification of physics is not necessarily a hindrance in our understanding of how the visual brain works, but it should be regarded as a new tool of neuroscience in discovering how humans see the world around them. More importantly, it raises new questions about the nature of art: is it merely an attempt to copy the real world around us or is it a more ambiguous, emotional interpretation of our environment?

Let's begin with a review on what we learned about ambiguous images and lateral inhibition during the lectures to understand the basis for many of the illusions discussed later in this paper. When looking at a contrast border between two different colors, the retinal output neurons are activated by both the light and dark areas of the light reflected from the image, as in the black and white checkerboard. Where the white and black squares touch, there is a contrast border, where the brightness of each field changes quickly. Signals from neurons picking up on the light from the contrast border are affected, or inhibited, by the signals from adjacent neurons. Approaching the darker border, neurons decrease in signal firing because of inhibition from the brighter border. Approaching the light border, signals increase because they are not completely inhibited by the darker areas (6). This physiology of the neurons in the retina explains several odd phenomena involving the changes in brightness along a contrast border, including the Hermann grid illusion, in which gray spots are seen in the spaces between the black squares of a grid (2). The lateral inhibition phenomenon is also related to the "filling in" of patterns and colors that the blind-spot misses where the retina meets the optic nerve (6).

So how does this relate to how we see artwork? Lateral inhibition is one reason why the brain interprets the stimuli from shadows and contrast borders as recognizable shapes and objects rather than an ambiguous collection of light and dark areas (2). As with the blind spot phenomenon, we tend to fill in fill in the gaps of ambiguous figures to form them into something that we can recognize based on our understanding of the real world. Humans are remarkably good at interpreting ambiguous collections of light and shadow into recognizable forms, particularly faces and bodies. Looking at only an ambiguous dark-red area on a green background in Figure 3 of the Nature article, our brains take the ambiguous stimuli and interpret it as the shadows of a man's face (1). The lighter red on the same green background is not as readily interpreted as a face, or at least a "shadow cue" that makes the image appear much more 3D and realistic. Figure 9 of the Nature article shows two dancers made up of only isolated patches of color, arranged in such a way to suggest two dancing figures (1). We do not look at such a drawing and see a collection of separate parts, but rather we attempt to put the shapes together to form a whole.

Other conventions and techniques of artists to convey certain illusions also reveal more about how we perceive images in art perhaps erroneously. One of the simplest and most common conventions in art is the line or contour drawing, usually the first step to creating any 2D piece of art. To create a drawing that is recognizable and discernable from the background of the piece, lines are used to trace a subject and separate it from the rest of the image. But in real life, we do not discern objects by a contour line separating the object from the "ground" or background, but rather our brains interpret brightness and shadow contrasts to allow us to recognize the separation between object and background (3).

The illusion of transparency also requires a form of lines crossing at certain points to create the impression of two layers passing in front of each other. These points where lines cross are called X-junctions, critical points at which the lines intersect to create the appearance of a transparent glass, for example. The X-junctions are all that are needed to create this illusion; in Figure 6 of the Nature article, the painting of the water glass is realistically transparent because of the multiple X-junctions that make up the rim and meniscus of the liquid, but there is no refractory distortion of the lemon slice within the glass as it would appear in real life (1). Yet this does nothing to diminish the illusion of a clear glass, once again indicating the simplified physics of our minds: even when not paying attention to impossible physics, we still perceive these images as transparent, or 3-dimensional, or whatever the illusion would lead us to see. Artwork, then, should not be looked at as simply an attempt to imitate life, for as we have seen with the illusions of contour lines, shadows and light, 2D paintings are not always true to life. Art is the subjective interpretation of the artist of how an object should look while still giving it credence of reality, like the illusion of X-junctions that approach, but do not match, transparency.

Other notable "mistakes" in artwork are not limited to the Renaissance pieces discussed earlier in the Nature article pointing out the inconsistent light and impossible shadows. Through personal observation, so many works of classic art do not follow the normal physics of lighting, yet these "errors" do nothing to detract from the overall perception or even of the piece as a whole. To use an example of one of my personal favorites, I cite Figure 1, a 19th century painting by William Bouguereau called Le Ravissement de Psyche; web-hosted by me (4). In this composition, the light is coming in from two different directions at once, both from behind the two figures as per the sunlight filtering behind them, but also from the front, as if both figures were being lit in a studio setting. This would be fine if the background lighting was consistent with the front lighting, but in the particular setting given, with mountains and a horizon, the figures would normally be in greater shadow if lighted only by the sunlight from behind. This sort of irregularity, however, is not immediately noticeable to the viewer, whose eyes are first attracted to the principal subject of the painting (the well-lit figures) and ignores the impossible lighting in lieu of contemplating the emotions that the painting conveys.

So what are the final implications of studying these "mistakes" and art's effect on the viewer? Expanding on the previously-mentioned idea of ambiguity, we turn to Impressionism, a style of art that relies on blurred suggestions and patterns that resolve themselves into recognizable shapes and figures. Interestingly, researchers have found certain neural connections in the amygdala and superior colliculus of the brain that respond more strongly to blurred or indistinct images, resulting in stronger emotional responses. For example, more areas of the amygdala are activated when looking at a blurry image of a fearful face, providing strong and quick fear-related emotions (5). The images of Impressionism work in much the same fashion; emotional centers of the brain respond more strongly to indistinct images. Therefore, is artwork simply a reaction to certain visual cues that activate the emotional brain centers? Given all that we know about the brain and consciousness, and after examining the strange ways in which our brains can be "fooled" into believing something is real when it isn't, I am inclined to agree. Creating artwork is a largely subjective process, and examining the myriad of ways in which art differs from (or imitates) real life can reveal much about the workings of the visual and creative brain.

References

1)Nature, 17 March: The Artist as Neuroscientist by Patrick Cavanagh, the starting point for my paper, provides many visual examples of artistic mistakes and how this relates to neurobiology.

2)The Psychology Papers, an interesting, if simplified, page about optical illusions and ambiguous images.

3)Picture Recognition in Animals and Humans (full article provided in PDF format) by Dalila Bovet and Jacques Vauclair, a very detailed paper about how humans and animals are both able to percieve pictoral images of real objects.

4)Figure 1, Le Ravissement de Psyche by William Bourguereau, simply an image that illustrates the subtle effects of impossible lighting in a painting. Uploaded to my own webhost.

5)Distinct spatial frequency sensitivities for processing faces and emotional expressions, by Patrik Vuilleumier et al., a paper describing the neural connections in the brain's emotional centers that respond most strongly to indistinct or fuzzed-out images.

6)Lateral Inhibition: documents from Serendip, a page from Serendip that clearly explains the process and implications of lateral inhibition in the retina's photoreceptors.


| Course Home | Serendip Home |